kokimashita
commited on
Commit
•
82df35d
1
Parent(s):
4c897b3
Create config.json
Browse files- config.json +34 -37
config.json
CHANGED
@@ -1,40 +1,37 @@
|
|
1 |
{
|
2 |
-
"
|
3 |
-
"
|
4 |
-
|
5 |
-
"bias": "none",
|
6 |
-
"fan_in_fan_out": false,
|
7 |
-
"inference_mode": true,
|
8 |
-
"init_lora_weights": true,
|
9 |
-
"layer_replication": null,
|
10 |
-
"layers_pattern": null,
|
11 |
-
"layers_to_transform": null,
|
12 |
-
"loftq_config": {},
|
13 |
-
"lora_alpha": 32,
|
14 |
-
"lora_dropout": 0.05,
|
15 |
-
"megatron_config": null,
|
16 |
-
"megatron_core": "megatron.core",
|
17 |
-
"modules_to_save": null,
|
18 |
-
"peft_type": "LORA",
|
19 |
-
"r": 16,
|
20 |
-
"rank_pattern": {},
|
21 |
-
"revision": null,
|
22 |
-
"target_modules": [
|
23 |
-
"o_proj",
|
24 |
-
"lm_head",
|
25 |
-
"up_proj",
|
26 |
-
"fc1",
|
27 |
-
"q_proj",
|
28 |
-
"linear_1",
|
29 |
-
"down_proj",
|
30 |
-
"out_proj",
|
31 |
-
"linear_2",
|
32 |
-
"fc2",
|
33 |
-
"gate_proj",
|
34 |
-
"k_proj",
|
35 |
-
"v_proj"
|
36 |
],
|
37 |
-
"
|
38 |
-
"
|
39 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "./checkpoints/vicuna-7b-v1-5",
|
3 |
+
"architectures": [
|
4 |
+
"LlamaForCausalLM"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
],
|
6 |
+
"bos_token_id": 1,
|
7 |
+
"eos_token_id": 2,
|
8 |
+
"hidden_act": "silu",
|
9 |
+
"hidden_size": 4096,
|
10 |
+
"image_aspect_ratio": "pad",
|
11 |
+
"initializer_range": 0.02,
|
12 |
+
"intermediate_size": 11008,
|
13 |
+
"max_position_embeddings": 4096,
|
14 |
+
"mm_hidden_size": 1024,
|
15 |
+
"mm_projector_lr": 2e-05,
|
16 |
+
"mm_projector_type": "mlp2x_gelu",
|
17 |
+
"mm_use_im_patch_token": false,
|
18 |
+
"mm_use_im_start_end": false,
|
19 |
+
"mm_vision_select_feature": "patch",
|
20 |
+
"mm_vision_select_layer": -2,
|
21 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
22 |
+
"model_type": "llava",
|
23 |
+
"num_attention_heads": 32,
|
24 |
+
"num_hidden_layers": 32,
|
25 |
+
"num_key_value_heads": 32,
|
26 |
+
"pad_token_id": 0,
|
27 |
+
"pretraining_tp": 1,
|
28 |
+
"rms_norm_eps": 1e-05,
|
29 |
+
"rope_scaling": null,
|
30 |
+
"tie_word_embeddings": false,
|
31 |
+
"torch_dtype": "float16",
|
32 |
+
"transformers_version": "4.31.0",
|
33 |
+
"tune_mm_mlp_adapter": false,
|
34 |
+
"use_cache": true,
|
35 |
+
"use_mm_proj": true,
|
36 |
+
"vocab_size": 32000
|
37 |
}
|